Bulg. J. Phys. vol.44 no.2 (2017), pp. 189-204



Development of Stochastic Daily Weather Generator Conditional on Atmospheric Circulation. Part 2: Daily Minimal and Maximal Temperature Models

N. Neykov, P. Neytchev
National Institute of Meteorology and Hydrology, Bulgarian Academy of Sciences, 66 Tsarigradsko shose Blvd., 1784 Sofia, Bulgaria
Abstract. Stochastic weather generator is developed to simulate synthetic time-series of daily precipitation totals, minimal and maximal temperatures. At the first step daily precipitation model is developed consisting of two components to describe the occurrence and intensity series, respectively. Binary logistic autoregression is used to fit the occurrence data, and the intensity series is fitted by gamma autoregressive model, conditional on atmospheric derivatives. At the second step the minimal and maximal daily temperatures are fitted to historical data by autoregressive models conditionally on precipitation occurrence and atmospheric derivatives. The daily time series from Kneja and Sadovo stations in Bulgaria are analyzed. Standard software for generalized linear models is used to perform the computations. Some potential difficulties are outlined.

Full-text: PDF

go back